Loss of independent control of joint movement in the impaired limb is a cardinal sign of stroke that is expressed in the form of stereotypic multi-joint movement patterns. Work from our laboratory has produced evidence for a loss of certain muscle coactivation patterns and associated abnormal shoulder/elbow torque coupling in the paretic arm of individuals with hemiparetic stroke. The general aim of this proposal is to elucidate the role of sensorimotor cortices in the loss of independent joint control of the paretic upper limb following stroke. We propose to study 20 individuals with chronic stroke and 20 control subjects to identify and quantify the characteristics of sensorimotor cortical activity related to the generation of isometric shoulder abduction/adduction and elbow flexion/extension torques. We hypothesize that an increased overlap of sensorimotor cortical activity is present in individuals with stroke (Aim 1). This aim will be realized using a 6 degree of freedom (DOF) load cell to provide simultaneous measurements of elbow and shoulder torques while measuring EEC from 163 scalp sites and EMG from 10 muscles in each arm. We postulate and have provided preliminary evidence for a significant overlap between cortical activity related to shoulder abduction and elbow flexion torque generation in chronic stroke subjects while not in able-bodied subjects. Subsequently, we plan to elucidate the relationship between the overlap of cortical activity and abnormal shoulder/elbow torque coupling in the paretic arm (Aim 2). We postulate that an increase in the overlap of cortical current activity is a reason for the abnormal torque coupling and serves as a predictor of the severity of discoordination in individuals with stroke. Finally, we propose to investigate the effect of a novel multidegree of freedom isometric training protocol on the overlap of cortical activity in individuals with stroke (Aim 3). Results from our recent work, using this training protocol, have shown a reduction in abnormal torque coupling between shoulder and elbow following an eight week intervention in all stroke subjects (n=7). We postulate and have provided preliminary evidence for a reduction in cortical-activity overlap in concurrence with a training-induced reduction in abnormal torque coupling. In conclusion, the combination of multichannel EEG imaging with a well controlled isometric elbow/shoulder torque generating task in subjects with chronic hemiparetic stroke is unmatched in the current literature. This study will provide new indices in the evaluation and study of discoordination as well as dynamic information on cortical plasticity in individuals with chronic hemiparetic stroke undergoing an upper extremity physical intervention. As such, the proposed study will substantially enhance our understanding of mechanisms driving discoordination in hemiparetic stroke and will provide foundations for further scientific inquiry into the development of more effective therapeutic interventions.